Personal Web Page for Simon Scott
I am a second-year graduate student in the Reconfigurable Computing research group at UC Berkeley. I did my undergraduate at the University of Cape Town in Electrical Engineering and Computer Science. I also have a M.Sc. in Electrical Engineering from the same university, for which I built a FPGA board for radar and telecommunications signal processing.
Previous projects on which I have worked include:
- VEGAS, a hybrid CPU/GPU/FPGA spectrometer for radio astronomy
- RHINO, an FPGA board for DSP for radar and telecommunications
- The acceleration of HMMER, a protein sequencing software tool, using FPGAs
- Image processing for synthetic-aperture radars using GPUs
I am currently working on eWallpaper, a type of electronic wallpaper in which a number of sensors and processing nodes are embedded. The motivation behind this concept is to remove the processing power from mobile devices and rather embed it within the environment itself. The first iteration of the eWallpaper will consist of a regular grid of 10 000 low-cost CPUs, each with its own radio transceiver, embedded directly into the wallpaper.
The large grid of radio transceivers in the wallpaper allows a number of interesting applications, such as radar imaging of a room and narrow-beam communications between between the wall and mobile devices within the room. Of particular interest to me is how to efficiently use a cluster of thousands of CPUs, connected together in a grid network. Although each node has limited processing and memory resources, all 10 000 CPUs together have substantial resources. Since many of the algorithms that need to implemented on the eWallpaper are embarrassingly-parallel (such as DSP and image processing), a message passing scheme may be a promising approach to building the massively-parallel system.
Click here to download my CV.
S. Cheng, M. Lin, H. J. Liu, S. Scott, and J. Wawrzynek, "Exploiting
memory-level parallelism in reconfigurable accelerators," in 2012 IEEE 20th
International Symposium on Field-Programmable Custom Computing Machines. 2012, pp. 157-160.
A. Bagula, M. Zennaro, G. Inggs, S. Scott, and D. Gascon, "Ubiquitous
sensor networking for development (USN4D): An application to pollution
monitoring," Sensors, vol. 12, pp. 391-414, Jan 2012.
S. Scott, "Rhino: Reconfigurable hardware interface for computation and
radio," Master's thesis, University of Cape Town, 2011.
M. Inggs, G. Inggs, A. Langman, and S. Scott, "Growing horns: Applying
the rhino software defined radio system to radar," in 2011 IEEE Radar
Conference (RADAR), May 2011, pp. 951-955.
A. Bagula, G. Inggs, S. Scott, and M. Zennaro, "Community sensor net-
works: An application to pollution maps," in Proceedings of the Interna-
tional Wireless Communication and Information Conference, Oct 2010.
M. Zennaro, A. Bagula, H. Ntareme, G. Inggs, and S. Scott, "On the rele-
vance of open wireless sensors for ngn," in Proceedings of the 2009 ITU-T
Kaleidoscope Conference: Innovations for Digital Inclusions, Aug 2009.
A. Bagula, G. Inggs, S. Scott, and M. Zennaro, "On the relevance of using
openwireless sensor networks in environment monitoring," Sensors, vol. 9,
pp. 4845-4868, Jun 2009.
Email: simonscott <at> berkeley <dot> edu
Phone: +1 510 501 2116
Mail: Berkeley Wireless Research Center
2108 Allston Way,
United States of America